CFIR may be useful when the project meets the following criteria:
(1) Your research question includes predicting and/or explaining implementation outcomes based on implementation determinants
(2) The unit of analysis is a defined Inner Setting that will be implementing and delivering the innovation, e.g., hospital, school, city
(3) The team has a methodologist and/or analyst with experience in implementation science methods and/or using CFIR
CFIR updates include:
In addition, a companion paper conceptualizing outcomes for use with CFIR was published (Damschroder, Reardon, Opra Widerquist, et al. 2022). These updates were based on user feedback and a full mapping of constructs from the original CFIR to the updated CFIR (as well as the rationale based on user feedback) is available in Additional File 5 in the updated CFIR manuscript (Damschroder, Reardon, Widerquist, et al. 2022). This mapping will be especially useful for teams that started projects using the original CFIR but want to present results using the updated CFIR.
The updated CFIR explicitly guides CFIR users to elicit “the degree to which” each construct manifests as defined. Perceptions and shared meanings, arising through social interactions among individuals in the workplace (Scroggins 2006), are an important influence on how people respond to this question, along with objective consideration of presence or absence of specific factors related to each construct. Thus, assessing the “degree to which” each construct manifests will likely elicit responses based on a blend of subjective judgements combined with objective fact; for example, Structural Characteristics: IT Infrastructure may capture the factual presence of an electronic health record system (EHR) as well as subjective perceptions about the degree to which that EHR supports functional performance in the Inner Setting.
The literature has recognized that the lack of a clean boundary between the innovation and implementation strategies is a contributor to implementation complexity (Butler et al. 2017); however, distinguishing between the innovation and implementation strategy is necessary for accurate attribution to implementation outcomes (Pinnock et al. 2017) and to identify appropriate areas for future intervention. As a result, the updated CFIR guides users to define the innovation (aka the “thing” (Curran 2020; Lengnick-Hall et al. 2022) being implemented), including the boundary between the innovation and implementation strategies. We encourage use of a reporting guideline to define the innovation (Albrecht et al. 2013; Butler et al. 2017; The AIMD Writing/Working Group et al. 2017; Hoffmann et al. 2014).
Rogers’ classic Diffusion of Innovation theory defines innovation as an idea, practice, or object that is perceived as new by an individual or other unit of adoption; if an idea seems new within a setting or for an individual, it is an innovation (E. Rogers 2003). This is a broad definition and includes any “thing” that is being implemented (Curran 2020).
While a clearly defined evidence-based innovation (EBI) is foundational in implementation science, and represents the most straightforward use of CFIR, the framework can be adapted to evaluate any “innovation.” For example, CFIR may help identify barriers and facilitators to increasing delivery of a previously implemented innovation, completing a quality improvement project, de-implementing an innovation, or using an implementation strategy. Using CFIR with non-EBI “things” is more challenging and requires additional effort when defining the domains in CFIR.
CFIR users have expressed difficulty operationalizing the Inner vs. Outer Setting in their projects and have given diverse recommendations from combining them into a single domain (because it was challenging to define a boundary between them) to separating each of them into multiple levels to account for multiple spheres of influence. For example, having national, regional, and local levels within the Outer Setting and departments, units, and project teams within the Inner Setting (Miake-Lye et al. 2020; Safaeinili et al. 2020).
While it can be challenging to differentiate between the Inner and Outer Settings, it is nonetheless important to define and delineate the boundary between the two domains for accurate attribution to implementation outcomes and to identify appropriate levels of future intervention. However, the updated CFIR states that there may be multiple Inner and Outer Settings as well as multiple levels within each Inner and Outer Setting; users can code at the levels most appropriate to their project.
Some users have questioned the inclusion of the Implementation Process Domain in CFIR because it appears to include strategies not determinants. As a determinant framework, CFIR includes determinants related to several spheres of influence: the innovation being implemented, the individuals involved, the settings, and the implementation process. The goal of this domain is to capture the use and quality of these implementation processes as determinants to implementation success, not to directly guide implementation.
CFIR is not designed to guide development of innovations. While many innovation development frameworks include assessing and understanding context (O’Cathain et al. 2019), we recommend selecting an innovation development framework for your project, e.g., Intervention Mapping (Fernandez, Ruiter, et al. 2019) or the Framework for Developing and Evaluating Complex Interventions (Skivington et al. 2021). Following selection of an innovation development framework, researchers may integrate CFIR into the framework to assess context.
CFIR is not a process model designed to guide the specific steps of implementation (Nilsen and Birken 2020); while CFIR includes an Implementation Process Domain, the goal of this domain is to capture the use and quality of these implementation processes as determinants to implementation success, not to directly guide implementation. While CFIR can guide assessment of potential barriers and facilitators to implementation, we recommend selecting a process model to guide implementation, e.g., Quality Implementation Framework (D. C. Meyers et al. 2012), Getting To Outcomes (Chinman et al. 2018), or Getting To Implementation (Rogal et al. 2020). Following selection of a process model, researchers may integrate CFIR into the model to assess context and identify potential barriers and facilitators to implementation.
Ideally, your chosen implementation outcome is one that is most proximal to the implementation strategy(s) being used and is an indicator of implementation (i.e., delivery of the innovation) within the Inner Setting.
Both qualitative (e.g., fidelity observations, interviews) and quantitative (e.g., surveys, administrative data) can be used to assess implementation outcomes. Questions that may be useful for assessing anticipated and/or actual implementation outcomes are included in Additional File 1; these questions must be customized to fit each project but can then be used as a part of data collection instruments. We recommend collecting an “objective” measure of implementation that is assessed by an outside evaluator, e.g., fidelity ratings or administrative data, that reflect the extent to which implementation is complete and equitable within and across Inner Settings.
Additional guidance and a mapping of implementation outcomes across RE-AIM (Glasgow et al. 2019) and the Implementation Outcomes Framework (E. Proctor et al. 2011) is available in the CFIR outcomes addendum (Damschroder, Reardon, Opra Widerquist, et al. 2022).
Innovation outcomes include the impact of the innovation on recipients, deliverers, and key-decision makers) (Damschroder, Reardon, Opra Widerquist, et al. 2022) and are innovation specific. While hybrid implementation-effectiveness designs are used to assess both implementation and innovation outcomes (Curran et al. 2012), measuring innovation effectiveness requires collecting additional data on innovation determinants (e.g., patient-level determinants) and outcomes (e.g., patient-level outcomes). See the CFIR outcomes addendum for more information (Damschroder, Reardon, Opra Widerquist, et al. 2022).
After identifying potential (or actual) barriers and facilitators to implementation using CFIR, strategies to mitigate barriers and leverage facilitators can be developed and/or identified via several participatory methods such as user-centered design (Knapp et al. 2022), Implementation Mapping (Walker et al. 2022), or concept mapping (Lewis et al. 2021). In addition, a tool to help users “match” strategies to barriers was developed using the original CFIR (Waltz et al. 2019), with strategies being drawn from Expert Recommendations for Implementing Change (ERIC) (Powell et al. 2015; Waltz et al. 2015).
Overall, you will use CFIR as described in this guide. You may find it useful to map components of the implementation strategy to constructs in the Implementation Process Domain, where some of the more common strategies are included as constructs. This will facilitate comparing how components of the implementation strategy appear in the data in each trial arm, i.e., how they manifest and/or interact differently with other constructs based on the implementation strategy used (Cannon et al. 2019).
When collecting data, researchers must be clear about the goal of data collection: 1) to predict and/or explain implementation outcomes based on implementation determinants (this is within the scope of CFIR); or 2) to predict and/or explain innovation outcomes based on innovation determinants (this is outside the scope of CFIR).
CFIR implementation determinants capture Inner Setting-level barriers and facilitators that predict and/or explain implementation outcomes, i.e., the innovation being implemented and delivered as intended in the Inner Setting. These determinants are denoted by the gray arrow in Figure 1 labeled CFIR Implementation Determinants. Data (qualitative and/or quantitative) on these determinants is best collected from individuals who have influence and/or power related to implementation (usually folks within the implementing setting); these typically include the key decision-makers and individuals implementing and/or delivering the innovation.
As a result, CFIR is not the appropriate framework to use when collecting data from recipients, unless recipients are also helping to implement and/or deliver the innovation in the Inner Setting. As reflected by Orlando et al., it is disappointing to note that “… while patients are part of the health-care organization and are essential to assessing intervention [innovation] effectiveness, they are a less influential component of implementation success in health-care settings than administrators and physicians” (emphases added) (Orlando et al. 2018). Although hospital systems are increasingly prioritizing patient-centered care, convening patient advisory boards, and involving patients in co-design of initiatives (Lyon et al. 2018; Dopp et al. 2019b), these efforts have not yet resulted in true power-sharing between innovation recipients and key decision-makers in the Inner Setting (Trofino 2003).
As a result, direct data collection from recipients does not usually inform implementation outcomes. Instead, data collection from key decision-makers and individuals implementing and/or delivering the innovation about their perceptions of recipients (e.g., recipient needs and characteristics), and how those perceptions encourage (or discourage) completing implementation, informs Implementation Outcomes. Although CFIR is often not appropriate for use with recipients (because they rarely hold roles as key decision-makers or innovation implementers/deliverers in the Inner Setting), we hope that will change. Recipients should have greater influence, authority, and power in systems; the updated CFIR highlighted the importance of implementation teams including innovation recipients (and innovation deliverers) as members. When recipients serve in that role, we strongly encourage using CFIR to collect data about implementation determinants from them – because they are also implementation team members. Ultimately, equitable population impact is only possible when recipients are integrally involved in implementation and all key constituencies share power and make decisions together.
In contrast to implementation determinants, innovation determinants capture recipient-level characteristics and/or experiences with the innovation that predict and/or explain innovation outcomes. These determinants are denoted by the gray arrow in Figure 1 labeled Innovation Determinants. Data (qualitative and/or quantitative) on these determinants is best collected from recipients. Innovation determinants include constructs or measures that are based on the theoretical framework underlying the innovation. For example, in a “small change” weight loss intervention designed for patients, innovation determinants included patient-level demographics, motivation and intention, and self-efficacy because the intervention was guided by social-psychological and goal-conflict theories (Lutes et al. 2013). This innovation was tested within a randomized clinical trial (Damschroder et al. 2014) and a subset of patient characteristics (innovation determinants) were explored in secondary analyses to help explain innovation outcomes (Masheb, Lutes, Kim, et al. 2015; Masheb, Lutes, Myra Kim, et al. 2015; Vimalananda et al. 2016; Janney et al. 2017). CFIR was not designed to capture these theory-derived determinants of innovation outcomes, and adapting CFIR constructs for this purpose separates them from the underlying organizational theory.